Well strictly speaking electromotive force in a galvanic cell must be, I think, positive because the EMF measured is a magnitude. So when it says "the electromotive force E of the above cell" it is referring to the EMF measured on the written cell, and EMF measured for any galvanic cell will be positive.
That does not mean that the thermodynamic concept of E with which we use electrode potential to consider Gibbs' free energy of reactions - represented by cell diagrams maybe but not in themselves cells, rather: reactions - cannot be both positive and negative, as with the Gibbs' free energy itself.
So, the bottom-line is: EMF measured, and electromotive force registered on any cell, is always going to be positive for any cell (providing it is not externally interfered with); it's only when looking at electrode potentials for reactions, thermodynamically, that we consider the direction and thus that we call the potential either negative or positive.
So I thought?